| Paper Abstract and Keywords |
| Presentation |
2006-09-01 14:30
A Method for Classifying Candidates of Tumor in PET/CT images and Its Application of Prognostic Observations
Shuhei Nitta, Hidekata Hontani (NIT), Tadanori Fukami, Tetsuya Yuasa, Takao Akatsuka (Yamagata Univ．), Jin Wu, Tohoru Takeda (Tsukuba Univ.), Noboru Oriuchi, Keigo Endo (Gunma Univ.), Yorihisa Watanabe (Yamagata Saisei Hospital) |
| Abstract |
(in Japanese) |
(See Japanese page) |
| (in English) |
In this article, we propose a method for classifying candidates of tumors detected in PET/CT images
and apply the proposed method to prognostic observations. Our method detects bright regions in a PET image as tumors. For improving the performance of the tumor detection, we select a proper threshold for binarizing the image according to location of a body. In addition, we extract a liver and kidneys that accumulate FDG using a probabilistic atlas, and we detect candidates of tumors in a liver region by a specialized method. Subsequently, we sort the detected tumor candidates by the degree of FDG accumulation. We make prognostic observations by
comparing a preoperative PET/CT image with apostoperative PET/CT image. This article shows results of the proposed method and discusses the performance. |
| Keyword |
(in Japanese) |
(See Japanese page) |
| (in English) |
PET/CT / probabilistic Atlas / tumor / prognostic observation / / / / |
| Reference Info. |
IEICE Tech. Rep., vol. 106, no. 225, MI2006-55, pp. 35-40, Sept. 2006. |
| Paper # |
MI2006-55 |
| Date of Issue |
2006-08-25 (MI) |
| ISSN |
Print edition: ISSN 0913-5685 |
| Download PDF |
|